Methionine (Met) dependence, the inability of cells to grow when Met is replaced in the culture medium by its immediate precursor homocysteine (Hcy) (Met-Hcy+medium), is a metabolic defect we have demonstrated that is found to be highly prevalent in and specific to cancer cells. In Met-Hcy+medium, human cancer cells that cannot maintain a high intracellular ratio of S-adenosylmethionine (AdoMet) to S-adenosylhomocysteine (AdoHcy) are Met dependent. Recent data from our laboratory indicate that, in human cancer cells the overall rate of transmethylation is elevated, but that certain substances such as DNA are undermethylated. Our results, therefore, indicate imbalanced transmethylation in human cancer cells and that Met dependence may be a result of the altered transmethylation. We have recently taken advantage of the metabolic defect of Met dependence to enable a sequential combination of adriamycin and vincristine to completely eliminate human Met-dependent cancer cells from a coculture with normal human fibroblasts which continued to proliferate during the treatment. Experiments will be performed to further understand the basis of unbalanced transmethylation in human cancer cell lines. Experiments in progress include the isolation of normal and cancer cell transmethylases by affinity chromotography with AdoHcy-agarose and separation of the affinity-purified proteins by a two-dimensional system comprising isolectric focusing and SDS polyacrylamide gel electrophoresis. Experiments are also in progress to determine the effects of methylation on the human rasHoncogene. A 6.4-kb rasH-recombinant is being methylated in vitro with the HpaII and HhaI methylases. In addition, the 2.9-kb SstI fragment of the rasHongogene, still containing all four known exons, has been cloned into the single-stranded DNA of bacteriophage M13 for the purpose of constructing a completely hemimethylated double-stranded hybrid rasH-M13 DNA in vitro. Both of the above methylated constructs will be compared to their nonmethylated counterparts with regard to their ability to oncogenically transform NIH 3T3 cells. Using rat monoclonal antibodies directed against the human rasHprotein p21 and the techniques of immunohistochemistry and immunoblotting, we are investigating the levels of p21 in Met-dependent cancer cell lines, Met-independent cancer cell variants selected from the former, and the multiple cell types of primary human tumor cells grown from individual explants of human tumors on collagen gels. The extent of methylation of the rasHgene will then be compared to the levels of rasHp21 and the oncogenicity of the tested cell types. (N)